BP神经网络算法对建筑保温材料性能的预测
|
摘要
针对建筑保温材料性能表征十分复杂、困难的情况,利用人工神经网络BP算法,建立了复合保温材料性能预测模型,模型由3层神经元组成,分别为输入层、隐含层和输出层。以炉渣复合材料性能与成分的关系为研究对象,采取108组实验数据对神经网络进行8 000次训练,神经网络输出值的平方平均误差为0.000 12。然后,选用18组实验数据对训练成熟的试验神经网络模型进行检测,并把检测样本的神经网络输出值和试验值进行比较。结果表明:所建立的网络能反映炉渣复合保温材料与材料性能之间的关系,为实验设计提供了新的思想,节省了时间和劳动力。
Faced with complexity and difficulty of performance of building thermal insulation materials,the composite insulation performance prediction model,which consisted of three neurons,including input layer,hidden layer and output layer,was made by BP Algorithm of artificial neural network. The relationship between the performance and composition of the slag composite was studied and 108 groups of experimental data were taken to train the neural network for 8 000 times,which finally drewa conclusion that the average error of the output value of the neural network was 0. 000 12. Then,18 groups of experimental data were used to test the neural network model,and the output values and test values of the test samples were compared. The results showthat the network can reflect the relationship between the thermal insulation material and the material properties,which provides a newidea for the experimental design,and saves time and labor.
Faced with complexity and difficulty of performance of building thermal insulation materials,the composite insulation performance prediction model,which consisted of three neurons,including input layer,hidden layer and output layer,was made by BP Algorithm of artificial neural network. The relationship between the performance and composition of the slag composite was studied and 108 groups of experimental data were taken to train the neural network for 8 000 times,which finally drewa conclusion that the average error of the output value of the neural network was 0. 000 12. Then,18 groups of experimental data were used to test the neural network model,and the output values and test values of the test samples were compared. The results showthat the network can reflect the relationship between the thermal insulation material and the material properties,which provides a newidea for the experimental design,and saves time and labor.
引文
[1]李亚辉,钱元弟.木质素纤维增强干混砂浆的制备与性能表征[J].建筑节能,2016,(4):45-48.
[2]李聚刚,陈晓龙,等.关键组分对高聚物乳液基酚醛树脂粉玻化微珠复合保温浆料性能的影响研究[J].建筑节能,2013,(10):37-41.
[3]王学成,李春珍.影响外墙外保温系统保温砂浆性能的因素[J].建筑节能,2010,(3):14-16.
[4]张训.无机保温砂浆对建筑能耗的影响[J].建筑节能,2012,(12):44-49.
[5]舒阳,杜宇平,等.高性能钢筋混凝土结构的耐久性研究[J].工程力学,2008,(S1):102-106.
[6]Taylor K K,Darscy J A.Prediction of the electronic properties of polymers using artificial neural netw orks[J].Polymer Preprints,2000,(1):331-332.
[7]Cherian R P,Smith L N,Midha P S.A neural network approach for selection of pow der metallurgy materials and process parameters[J].Artificial Intelligence in Engineering,2000,(14):39-44.
[8]舒阳,舒子杨,王敬东,等.关于民用建筑墙体自保温的研究[C]//第21届全国结构工程学术会议论文集:第Ⅲ册.北京:工程力学出版社,2012.
[9]贾阁森,严捍东,黄国晖.基于BP神经网络的橡胶粉砂浆性能的预测[J].混凝土,2008,(5):109-111.
[10]舒阳,王敬东,徐禄吉,等.高性能钢筋混凝土结构耐久性智能评估系统研究[J].混凝土,2008,(1):118-121.
[11]舒阳.盐渍土质中钢筋混凝土桩的耐久性应用研究[D].天津:天津大学,2007.
[12]尹海莲,胡自立.基于BP神经网络的复合材料性能预测[J].南京航空航天大学学报,2006,(2):234-238.
[2]李聚刚,陈晓龙,等.关键组分对高聚物乳液基酚醛树脂粉玻化微珠复合保温浆料性能的影响研究[J].建筑节能,2013,(10):37-41.
[3]王学成,李春珍.影响外墙外保温系统保温砂浆性能的因素[J].建筑节能,2010,(3):14-16.
[4]张训.无机保温砂浆对建筑能耗的影响[J].建筑节能,2012,(12):44-49.
[5]舒阳,杜宇平,等.高性能钢筋混凝土结构的耐久性研究[J].工程力学,2008,(S1):102-106.
[6]Taylor K K,Darscy J A.Prediction of the electronic properties of polymers using artificial neural netw orks[J].Polymer Preprints,2000,(1):331-332.
[7]Cherian R P,Smith L N,Midha P S.A neural network approach for selection of pow der metallurgy materials and process parameters[J].Artificial Intelligence in Engineering,2000,(14):39-44.
[8]舒阳,舒子杨,王敬东,等.关于民用建筑墙体自保温的研究[C]//第21届全国结构工程学术会议论文集:第Ⅲ册.北京:工程力学出版社,2012.
[9]贾阁森,严捍东,黄国晖.基于BP神经网络的橡胶粉砂浆性能的预测[J].混凝土,2008,(5):109-111.
[10]舒阳,王敬东,徐禄吉,等.高性能钢筋混凝土结构耐久性智能评估系统研究[J].混凝土,2008,(1):118-121.
[11]舒阳.盐渍土质中钢筋混凝土桩的耐久性应用研究[D].天津:天津大学,2007.
[12]尹海莲,胡自立.基于BP神经网络的复合材料性能预测[J].南京航空航天大学学报,2006,(2):234-238.